%matplotlib inline
import xarray as xr
import pandas as pd
import numpy as np
from cads_api_client import ApiClient
import os
from pathlib import Path
import matplotlib.pyplot as plt
import matplotlib.dates as mdates
import seaborn as sns
import Visualization_functions as vf
import importlib
importlib.reload(vf)
import psutil

process = psutil.Process()
print(f"Memory usage: {process.memory_info().rss / 1024 ** 2:.2f} MB")

Memory usage: 50702.56 MB

variable = "total_column_water_vapour_era5"  # zenith_total_delay or total_column_water_vapour or total_column_water_vapour_era5
starting_year, ending_year = 1996, 2014
network_type = "epn_repro2"  # igs_repro3 or igs_daily or epn_repro2

custom_base = input(
    "Enter base directory to save files (leave empty to use current folder): "
).strip()
base_dir = Path(custom_base) if custom_base else Path.cwd()
base_dir = base_dir / "Datas" / "EGU2025"  # /home/hbreton/Documents

client = ApiClient()
dataset = "insitu-observations-gnss"

[2024-06-16T00:00:00] CDS API syntax is changed and some keys or parameter names may have also changed. To avoid requests failing, please use the "Show API request code" tool on the dataset Download Form to check you are using the correct syntax for your API request.

for year in range(starting_year, ending_year + 1):
    year_str = str(year)
    request = {
        "network_type": network_type,
        "variable": [variable],
        "year": year_str,
        "month": [f"{m:02d}" for m in range(1, 13)],
        "day": [f"{d:02d}" for d in range(1, 32)],
        "format": "netcdf",
    }
    file_path = f"{base_dir}/{year_str}/{variable}_{network_type}_{year_str}.nc"

    if os.path.exists(file_path):
        print(f"File already exists: {file_path}. Skipping download.")
    else:
        try:
            os.makedirs(os.path.dirname(file_path), exist_ok=True)
            client.retrieve(dataset, **request, target=file_path)
            print(f"Download successful for {year_str}")
        except Exception as e:
            print(f"Error during download for {year_str}: {e}")

f38f961ce993608eebbcffa781f653c0.nc:   0%|          | 0.00/2.06M [00:00<?, ?B/s]

Download successful for 1996

4741e109ad11ab7e32d1c2d4fd3c3289.nc:   0%|          | 0.00/3.04M [00:00<?, ?B/s]

Download successful for 1997

740528945ccde586465bf283063e90a6.nc:   0%|          | 0.00/3.99M [00:00<?, ?B/s]

Download successful for 1998

e65ecb75c5f96ce2bad2f939b63b4ad3.nc:   0%|          | 0.00/4.51M [00:00<?, ?B/s]

Download successful for 1999

c6c987b6f52aa01a6157341f44549bff.nc:   0%|          | 0.00/5.45M [00:00<?, ?B/s]

Download successful for 2000

datasets = []
for year in range(starting_year, ending_year + 1):
    file_path = f"/{base_dir}/{year}/{variable}_{network_type}_{year}.nc"
    if os.path.exists(file_path):
        cadsobs_dataset = xr.open_dataset(file_path, engine="netcdf4")
        cadsobs_dataset.observed_variable.to_index().unique()
        cadsobs_dataset[
            "observed_variable"
        ] = cadsobs_dataset.observed_variable.str.decode("UTF-8")
        df_cadsobs_dataset = (
            cadsobs_dataset.to_dataframe()
            .set_index(["primary_station_id", "observed_variable"], inplace=False)
            .sort_index(level="primary_station_id")
        )
        datasets.append(df_cadsobs_dataset)
        print(f"Dataset loaded for {year}: {file_path}")
    else:
        print(f"File not found: {file_path}")

Dataset loaded for 1996: //home/hbreton/Documents/Datas/EGU2025/1996/total_column_water_vapour_era5_epn_repro2_1996.nc
Dataset loaded for 1997: //home/hbreton/Documents/Datas/EGU2025/1997/total_column_water_vapour_era5_epn_repro2_1997.nc
Dataset loaded for 1998: //home/hbreton/Documents/Datas/EGU2025/1998/total_column_water_vapour_era5_epn_repro2_1998.nc
Dataset loaded for 1999: //home/hbreton/Documents/Datas/EGU2025/1999/total_column_water_vapour_era5_epn_repro2_1999.nc
Dataset loaded for 2000: //home/hbreton/Documents/Datas/EGU2025/2000/total_column_water_vapour_era5_epn_repro2_2000.nc
Dataset loaded for 2001: //home/hbreton/Documents/Datas/EGU2025/2001/total_column_water_vapour_era5_epn_repro2_2001.nc
Dataset loaded for 2002: //home/hbreton/Documents/Datas/EGU2025/2002/total_column_water_vapour_era5_epn_repro2_2002.nc
Dataset loaded for 2003: //home/hbreton/Documents/Datas/EGU2025/2003/total_column_water_vapour_era5_epn_repro2_2003.nc
Dataset loaded for 2004: //home/hbreton/Documents/Datas/EGU2025/2004/total_column_water_vapour_era5_epn_repro2_2004.nc
Dataset loaded for 2005: //home/hbreton/Documents/Datas/EGU2025/2005/total_column_water_vapour_era5_epn_repro2_2005.nc
Dataset loaded for 2006: //home/hbreton/Documents/Datas/EGU2025/2006/total_column_water_vapour_era5_epn_repro2_2006.nc
Dataset loaded for 2007: //home/hbreton/Documents/Datas/EGU2025/2007/total_column_water_vapour_era5_epn_repro2_2007.nc
Dataset loaded for 2008: //home/hbreton/Documents/Datas/EGU2025/2008/total_column_water_vapour_era5_epn_repro2_2008.nc
Dataset loaded for 2009: //home/hbreton/Documents/Datas/EGU2025/2009/total_column_water_vapour_era5_epn_repro2_2009.nc
Dataset loaded for 2010: //home/hbreton/Documents/Datas/EGU2025/2010/total_column_water_vapour_era5_epn_repro2_2010.nc
Dataset loaded for 2011: //home/hbreton/Documents/Datas/EGU2025/2011/total_column_water_vapour_era5_epn_repro2_2011.nc
Dataset loaded for 2012: //home/hbreton/Documents/Datas/EGU2025/2012/total_column_water_vapour_era5_epn_repro2_2012.nc
Dataset loaded for 2013: //home/hbreton/Documents/Datas/EGU2025/2013/total_column_water_vapour_era5_epn_repro2_2013.nc
Dataset loaded for 2014: //home/hbreton/Documents/Datas/EGU2025/2014/total_column_water_vapour_era5_epn_repro2_2014.nc

df = pd.concat(datasets, axis=0)
output_dir = base_dir / "full_time_series_data"
output_dir.mkdir(parents=True, exist_ok=True)
df.to_parquet(
    f"{base_dir}/full_time_series_data/{variable}_{network_type}_{starting_year}_{ending_year}.parquet"
)

del datasets

custom_base = input(
    "Enter base directory where you saves your files (leave empty to use current folder): "
).strip()
base_dir = Path(custom_base) if custom_base else Path.cwd()
base_dir = base_dir / "Datas" / "EGU2025"  # /home/hbreton/Documents

starting_year, ending_year = 2000, 2019 #adapt to your data

df_ZTD_igs_repro3 = pd.read_parquet(
    f"{base_dir}/full_time_series_data/zenith_total_delay_igs_repro3_{starting_year}_{ending_year}.parquet"
)
df_ZTD_igs_repro3 = df_ZTD_igs_repro3.reset_index()
df_ZTD_igs_repro3["primary_station_id"] = df_ZTD_igs_repro3[
    "primary_station_id"
].astype(str)
df_ZTD_igs_repro3 = df_ZTD_igs_repro3.set_index(
    ["primary_station_id", "observed_variable"], inplace=False
).sort_index(level="primary_station_id")

starting_year, ending_year = 2000, 2019 #adapt to your data

df_PWC_igs_repro3 = pd.read_parquet(
    f"{base_dir}/full_time_series_data/total_column_water_vapour_igs_repro3_{starting_year}_{ending_year}.parquet"
)
df_PWC_igs_repro3 = df_PWC_igs_repro3.reset_index()
df_PWC_igs_repro3["primary_station_id"] = df_PWC_igs_repro3[
    "primary_station_id"
].astype(str)
df_PWC_igs_repro3 = df_PWC_igs_repro3.set_index(
    ["primary_station_id", "observed_variable"], inplace=False
).sort_index(level="primary_station_id")

starting_year, ending_year = 2000, 2019 #adapt to your data

df_PWC_era5_igs_repro3 = pd.read_parquet(
    f"{base_dir}/full_time_series_data/total_column_water_vapour_era5_igs_repro3_{starting_year}_{ending_year}.parquet"
)
df_PWC_era5_igs_repro3 = df_PWC_era5_igs_repro3.reset_index()
df_PWC_era5_igs_repro3["primary_station_id"] = df_PWC_era5_igs_repro3[
    "primary_station_id"
].astype(str)
df_PWC_era5_igs_repro3 = df_PWC_era5_igs_repro3.set_index(
    ["primary_station_id", "observed_variable"], inplace=False
).sort_index(level="primary_station_id")

df_PWC_igs_repro3

starting_year, ending_year = 2000, 2023

df_PWC_igs_daily = pd.read_parquet(
    f"{base_dir}/full_time_series_data/total_column_water_vapour_igs_daily_{starting_year}_{ending_year}.parquet"
)
df_PWC_igs_daily = df_PWC_igs_daily.reset_index()
df_PWC_igs_daily["primary_station_id"] = df_PWC_igs_daily["primary_station_id"].astype(
    str
)
df_PWC_igs_daily = df_PWC_igs_daily.set_index(
    ["primary_station_id", "observed_variable"], inplace=False
).sort_index(level="primary_station_id")

starting_year, ending_year = 2000, 2023

df_ZTD_igs_daily = pd.read_parquet(
    f"{base_dir}/full_time_series_data/zenith_total_delay_igs_daily_{starting_year}_{ending_year}.parquet"
)
df_ZTD_igs_daily = df_ZTD_igs_daily.reset_index()
df_ZTD_igs_daily["primary_station_id"] = df_ZTD_igs_daily["primary_station_id"].astype(
    str
)
df_ZTD_igs_daily = df_ZTD_igs_daily.set_index(
    ["primary_station_id", "observed_variable"], inplace=False
).sort_index(level="primary_station_id")

starting_year, ending_year = 2000, 2023

df_PWC_era5_igs_daily = pd.read_parquet(
    f"{base_dir}/full_time_series_data/total_column_water_vapour_era5_igs_daily_{starting_year}_{ending_year}.parquet"
)
df_PWC_era5_igs_daily = df_PWC_era5_igs_daily.reset_index()
df_PWC_era5_igs_daily["primary_station_id"] = df_PWC_era5_igs_daily[
    "primary_station_id"
].astype(str)
df_PWC_era5_igs_daily = df_PWC_era5_igs_daily.set_index(
    ["primary_station_id", "observed_variable"], inplace=False
).sort_index(level="primary_station_id")

df_PWC_igs_daily

starting_year, ending_year = 1996, 2014 #adapt to your data

df_ZTD_epn_repro2 = pd.read_parquet(
    f"{base_dir}/full_time_series_data/zenith_total_delay_epn_repro2_{starting_year}_{ending_year}.parquet"
)
df_ZTD_epn_repro2 = df_ZTD_epn_repro2.reset_index()
df_ZTD_epn_repro2["primary_station_id"] = df_ZTD_epn_repro2[
    "primary_station_id"
].astype(str)
df_ZTD_epn_repro2 = df_ZTD_epn_repro2.set_index(
    ["primary_station_id", "observed_variable"], inplace=False
).sort_index(level="primary_station_id")

starting_year, ending_year = 1996, 2014 #adapt to your data

df_PWC_epn_repro2 = pd.read_parquet(
    f"{base_dir}/full_time_series_data/total_column_water_vapour_epn_repro2_{starting_year}_{ending_year}.parquet"
)
df_PWC_epn_repro2 = df_PWC_epn_repro2.reset_index()
df_PWC_epn_repro2["primary_station_id"] = df_PWC_epn_repro2[
    "primary_station_id"
].astype(str)
df_PWC_epn_repro2 = df_PWC_epn_repro2.set_index(
    ["primary_station_id", "observed_variable"], inplace=False
).sort_index(level="primary_station_id")

starting_year, ending_year = 1996, 2014 #adapt to your data

df_PWC_era5_epn_repro2 = pd.read_parquet(
    f"{base_dir}/full_time_series_data/total_column_water_vapour_era5_epn_repro2_{starting_year}_{ending_year}.parquet"
)
df_PWC_era5_epn_repro2 = df_PWC_era5_epn_repro2.reset_index()
df_PWC_era5_epn_repro2["primary_station_id"] = df_PWC_era5_epn_repro2[
    "primary_station_id"
].astype(str)
df_PWC_era5_epn_repro2 = df_PWC_era5_epn_repro2.set_index(
    ["primary_station_id", "observed_variable"], inplace=False
).sort_index(level="primary_station_id")

df_PWC_epn_repro2

repro3_stations = df_PWC_igs_repro3.index.get_level_values(
    "primary_station_id"
).nunique()

daily_stations = df_PWC_igs_daily.index.get_level_values(
    "primary_station_id"
).nunique()

repro2_stations = df_PWC_epn_repro2.index.get_level_values(
    "primary_station_id"
).nunique()

repro3_station_list = df_PWC_igs_repro3.index.get_level_values(
    "primary_station_id"
).unique()

daily_station_list = df_PWC_igs_daily.index.get_level_values(
    "primary_station_id"
).unique()

repro2_station_list = df_PWC_epn_repro2.index.get_level_values(
    "primary_station_id"
).unique()


common_stations_repro3_daily = set(repro3_station_list).intersection(set(daily_station_list))
common_stations_repro3_repro2 = set(repro3_station_list).intersection(set(repro2_station_list))
common_stations_daily_repro2 = set(daily_station_list).intersection(set(repro2_station_list))

print(f"Number of unique stations in igs_daily: {daily_stations}")
print(f"Number of unique stations in igs_repro3: {repro3_stations}")
print(f"Number of unique stations in epn_repro2: {repro2_stations} \n")

print(f"Number of common stations between igs_repro3 and igs_daily: {len(common_stations_repro3_daily)}")
print(f"Number of common stations between igs_daily and epn_repro2: {len(common_stations_daily_repro2)}")
print(f"Number of common stations between igs_repro3 and epn_repro2: {len(common_stations_repro3_repro2)}")

Number of unique stations in igs_daily: 682
Number of unique stations in igs_repro3: 422
Number of unique stations in epn_repro2: 305 

Number of common stations between igs_repro3 and igs_daily: 422
Number of common stations between igs_daily and epn_repro2: 105
Number of common stations between igs_repro3 and epn_repro2: 84

vf.plot_time_series(df_ZTD_igs_repro3, "MATE00ITA", "zenith_total_delay", "IGSrepro3")
vf.plot_time_series(df_ZTD_epn_repro2, "MATE00ITA", "zenith_total_delay", "EPNrepro2")
vf.plot_time_series(df_PWC_igs_daily, "FORT00BRA", "precipitable_water_column", "IGSdaily")
vf.plot_time_series(df_PWC_era5_igs_repro3, "MATE00ITA", "precipitable_water_column_era5", "IGSrepro3")

try:
    df_equipment = pd.read_csv(
        "/home/hbreton/Documents/Datas/IGS/IGS.eq_change_list.RA.debfin.dates.csv" # Change this path to your file
    )
    df_equipment["Change_Date"] = pd.to_datetime(
        df_equipment["Change_Date"]
    ).dt.strftime("%Y-%m-%d %H:%M:%S")

except FileNotFoundError:
    # Create an empty dataframe with the same structure if file doesn't exist
    df_equipment = pd.DataFrame(
        {"Station": [], "Year": [], "Doy": [], "Change_Date": [], "Type": []}
    )
    print("Equipment file not found, creation of an empty dataframe")

df_equipment #ensure your file got the right format

df_repro3 = df_ZTD_igs_repro3.reset_index()[
    [
        "primary_station_id",
        "report_timestamp",
        "observation_value",
        "uncertainty_value1",
    ]
]
df_daily = df_ZTD_igs_daily.reset_index()[
    [
        "primary_station_id",
        "report_timestamp",
        "observation_value",
        "uncertainty_value1",
    ]
]

df_diff_daily_repro3 = pd.merge(
    df_daily,
    df_repro3,
    on=["primary_station_id", "report_timestamp"],
    how="inner",
    suffixes=("_daily", "_repro3"),
)

del df_daily
del df_repro3

df_diff_daily_repro3["ZTD_diff"] = (
    df_diff_daily_repro3["observation_value_daily"]
    - df_diff_daily_repro3["observation_value_repro3"]
)
df_diff_daily_repro3 = df_diff_daily_repro3.rename(
    columns={
        "primary_station_id": "Station",
        "observation_value_daily": "ZTD_daily",
        "observation_value_repro3": "ZTD_repro3",
        "report_timestamp": "Date",
        "uncertainty_value1_daily": "sigma_ZTD_daily",
        "uncertainty_value1_repro3": "sigma_ZTD_repro3",
    }
)
df_diff_daily_repro3 = df_diff_daily_repro3.set_index("Station")

df_diff_daily_repro3.to_parquet(
    f"{base_dir}/full_time_series_data/diff_zenith_total_delay_igs_daily_igs_repro3.parquet"
)

df_repro3 = df_PWC_igs_repro3.reset_index()[
    [
        "primary_station_id",
        "report_timestamp",
        "observation_value",
        "uncertainty_value5",
    ]
]
df_daily = df_PWC_igs_daily.reset_index()[
    [
        "primary_station_id",
        "report_timestamp",
        "observation_value",
        "uncertainty_value5",
    ]
]

df_diff_daily_repro3 = pd.merge(
    df_daily,
    df_repro3,
    on=["primary_station_id", "report_timestamp"],
    how="inner",
    suffixes=("_daily", "_repro3"),
)
del df_daily
del df_repro3

df_diff_daily_repro3["PWC_diff"] = (
    df_diff_daily_repro3["observation_value_daily"]
    - df_diff_daily_repro3["observation_value_repro3"]
)
df_diff_daily_repro3 = df_diff_daily_repro3.rename(
    columns={
        "primary_station_id": "Station",
        "observation_value_daily": "PWC_daily",
        "observation_value_repro3": "PWC_repro3",
        "report_timestamp": "Date",
        "uncertainty_value5_daily": "sigma_PWC_daily",
        "uncertainty_value5_repro3": "sigma_PWC_repro3",
    }
)
df_diff_daily_repro3 = df_diff_daily_repro3.set_index("Station")

df_diff_daily_repro3.to_parquet(
    f"{base_dir}/full_time_series_data/diff_total_column_water_vapour_igs_daily_igs_repro3.parquet"
)

df_repro3 = df_PWC_igs_repro3.reset_index()[
    ["primary_station_id", "report_timestamp", "observation_value"]
]
df_era5 = df_PWC_era5_igs_repro3.reset_index()[
    ["primary_station_id", "report_timestamp", "observation_value"]
]

df_diff_repro3_era5 = pd.merge(
    df_repro3,
    df_era5,
    on=["primary_station_id", "report_timestamp"],
    how="inner",
    suffixes=("_repro3", "_era5"),
)
del df_repro3
del df_era5

df_diff_repro3_era5["PCW_diff"] = (
    df_diff_repro3_era5["observation_value_repro3"]
    - df_diff_repro3_era5["observation_value_era5"]
)
df_diff_repro3_era5 = df_diff_repro3_era5.rename(
    columns={
        "primary_station_id": "Station",
        "observation_value_repro3": "PCW_repro3",
        "observation_value_era5": "PCW_era5",
        "report_timestamp": "Date",
    }
)
df_diff_repro3_era5 = df_diff_repro3_era5.set_index("Station")

df_diff_repro3_era5.to_parquet(
    f"{base_dir}/full_time_series_data/diff_total_column_water_vapour_igs_repro3_era5_{starting_year}_{ending_year}.parquet"
)

df_daily = df_PWC_igs_daily.reset_index()[
    ["primary_station_id", "report_timestamp", "observation_value"]
]
df_era5 = df_PWC_era5_igs_daily.reset_index()[
    ["primary_station_id", "report_timestamp", "observation_value"]
]

df_diff_daily_era5 = pd.merge(
    df_daily,
    df_era5,
    on=["primary_station_id", "report_timestamp"],
    how="inner",
    suffixes=("_daily", "_era5"),
)
del df_daily
del df_era5

df_diff_daily_era5["PCW_diff"] = (
    df_diff_daily_era5["observation_value_daily"]
    - df_diff_daily_era5["observation_value_era5"]
)
df_diff_daily_era5 = df_diff_daily_era5.rename(
    columns={
        "primary_station_id": "Station",
        "observation_value_daily": "PCW_daily",
        "observation_value_era5": "PCW_era5",
        "report_timestamp": "Date",
    }
)
df_diff_daily_era5 = df_diff_daily_era5.set_index("Station")

df_diff_daily_era5.to_parquet(
    f"{base_dir}/full_time_series_data/diff_total_column_water_vapour_igs_daily_era5_{starting_year}_{ending_year}.parquet"
)

df_diff_ztd_daily_repro3 = pd.read_parquet(
    f"{base_dir}/full_time_series_data/diff_zenith_total_delay_igs_daily_igs_repro3.parquet"
)
df_diff_pwc_daily_repro3 = pd.read_parquet(
    f"{base_dir}/full_time_series_data/diff_total_column_water_vapour_igs_daily_igs_repro3.parquet"
)
df_diff_pwc_repro3_era5 = pd.read_parquet(
    f"{base_dir}/full_time_series_data/diff_total_column_water_vapour_igs_repro3_era5_2000_2019.parquet"
)
df_diff_pwc_daily_era5 = pd.read_parquet(
    f"{base_dir}/full_time_series_data/diff_total_column_water_vapour_igs_daily_era5_2000_2023.parquet"
)

dfs = {
    "ΔZTD[IGSdaily - IGSrepro3]": df_diff_ztd_daily_repro3,
    "ΔPWC[IGSdaily - ERA5]": df_diff_pwc_daily_era5,
    "ΔPWC[IGSrepro3 - ERA5]": df_diff_pwc_repro3_era5,
    "ΔPWC[IGSdaily - IGSrepro3]": df_diff_pwc_daily_repro3,
}

for i, (name, df) in enumerate(dfs.items()):
    counts = df.groupby("Station").size()
    counts_sorted = counts.sort_values(ascending=True)
    low_data_stations = counts_sorted[counts_sorted < 26280].index.tolist()

    print(f"Stations with less than 26280 days: {len(low_data_stations)}")
    if len(low_data_stations) > 0:
        print(f"Example stations with low data: {low_data_stations[:5]}")

    if "low_data_stations_dict" not in globals():
        globals()["low_data_stations_dict"] = {}

    globals()["low_data_stations_dict"][name] = low_data_stations

    if len(low_data_stations) > 0:
        df = dfs[name]
        df_filtered = df[~df.index.get_level_values("Station").isin(low_data_stations)]
        filtered_df_name = f"df_diff_{name.replace(' - ', '_')}_filtered"
        globals()[filtered_df_name] = df_filtered
        removed_rows = len(df) - len(df_filtered)
        print(
            f"\nFiltered out {removed_rows} rows ({removed_rows/len(df)*100:.2f}%) from {len(low_data_stations)} stations"
        )
        print(
            f"Original dataframe: {len(df)} rows, Filtered dataframe: {len(df_filtered)} rows"
        )

print("Saving filtered dataframes to parquet files...")
pwc_daily_era5_filtered = globals()["df_diff_ΔPWC[IGSdaily_ERA5]_filtered"]
pwc_repro3_era5_filtered = globals()["df_diff_ΔPWC[IGSrepro3_ERA5]_filtered"]
pwc_daily_repro3_filtered = globals()["df_diff_ΔPWC[IGSdaily_IGSrepro3]_filtered"]
ztd_daily_repro3_filtered = globals()["df_diff_ΔZTD[IGSdaily_IGSrepro3]_filtered"]

pwc_daily_era5_filtered.to_parquet(
    f"{base_dir}/full_time_series_data/Over_3_years_diff_total_column_water_vapour_igs_daily_era5_2000_2023.parquet"
)
pwc_repro3_era5_filtered.to_parquet(
    f"{base_dir}/full_time_series_data/Over_3_years_diff_total_column_water_vapour_igs_repro3_era5_2000_2019.parquet"
)
pwc_daily_repro3_filtered.to_parquet(
    f"{base_dir}/full_time_series_data/Over_3_years_diff_total_column_water_vapour_igs_daily_igs_repro3.parquet"
)
ztd_daily_repro3_filtered.to_parquet(
    f"{base_dir}/full_time_series_data/Over_3_years_diff_zenith_total_delay_igs_daily_igs_repro3.parquet"
)

print("All filtered dataframes saved successfully.")

Stations with less than 26280 days: 113
Example stations with low data: ['FAIV00USA', 'TABV00USA', 'KWJ100MHL', 'SUNM00AUS', 'OBET00DEU']

Filtered out 1489728 rows (2.28%) from 113 stations
Original dataframe: 65283864 rows, Filtered dataframe: 63794136 rows
Stations with less than 26280 days: 17
Example stations with low data: ['MTBG00AUT', 'FREE00BHS', 'KWJ100MHL', 'TGCV00CPV', 'MDVO00RUS']

Filtered out 309532 rows (0.68%) from 17 stations
Original dataframe: 45666183 rows, Filtered dataframe: 45356651 rows
Stations with less than 26280 days: 68
Example stations with low data: ['MTBG00AUT', 'BOAV00BRA', 'WTZ300DEU', 'KWJ100MHL', 'SUNM00AUS']

Filtered out 934346 rows (2.49%) from 68 stations
Original dataframe: 37479676 rows, Filtered dataframe: 36545330 rows
Stations with less than 26280 days: 68
Example stations with low data: ['MTBG00AUT', 'BOAV00BRA', 'WTZ300DEU', 'KWJ100MHL', 'SUNM00AUS']

Filtered out 934346 rows (2.49%) from 68 stations
Original dataframe: 37479676 rows, Filtered dataframe: 36545330 rows
Saving filtered dataframes to parquet files...
All filtered dataframes saved successfully.

pd.DataFrame({'Station': [], 'ZTD_diff': []}).set_index('Station')

pd.DataFrame({'Station': [], 'PWC_diff': []}).set_index('Station')

df_diff_ztd_daily_repro3 = pd.read_parquet(
    f"{base_dir}/full_time_series_data/Over_3_years_diff_zenith_total_delay_igs_daily_igs_repro3.parquet"
)
df_diff_ztd_daily_repro2 = pd.DataFrame({'Station': [], 'ZTD_diff': []}).set_index('Station')
df_diff_ztd_repro3_repro2 = pd.DataFrame({'Station': [], 'ZTD_diff': []}).set_index('Station')



df_diff_pwc_daily_repro3 = pd.read_parquet(
    f"{base_dir}/full_time_series_data/Over_3_years_diff_total_column_water_vapour_igs_daily_igs_repro3.parquet"
)
df_diff_pwc_daily_repro2 = pd.DataFrame({'Station': [], 'PWC_diff': []}).set_index('Station')
df_diff_pwc_repro3_repro2 = pd.DataFrame({'Station': [], 'PWC_diff': []}).set_index('Station')



df_diff_pwc_repro3_era5 = pd.read_parquet(
    f"{base_dir}/full_time_series_data/Over_3_years_diff_total_column_water_vapour_igs_repro3_era5_2000_2019.parquet"
)
df_diff_pwc_daily_era5 = pd.read_parquet(
    f"{base_dir}/full_time_series_data/Over_3_years_diff_total_column_water_vapour_igs_daily_era5_2000_2023.parquet"
)
df_diff_pwc_repro2_era5 = pd.DataFrame({'Station': [], 'PWC_diff': []}).set_index('Station')

dataframes_dict = {
    "ΔZTD[IGSdaily - IGSrepro3]": df_diff_ztd_daily_repro3,
    "ΔZTD[IGSdaily - EPNrepro2]": df_diff_ztd_daily_repro2,
    "ΔZTD[IGSrepro3 - EPNrepro2]": df_diff_ztd_repro3_repro2,

    "ΔPWC[IGSdaily - IGSrepro3]": df_diff_pwc_daily_repro3,
    "ΔPWC[IGSdaily - EPNrepro2]": df_diff_pwc_daily_repro2,
    "ΔPWC[IGSrepro3 - EPNrepro2]": df_diff_pwc_repro3_repro2,

    "ΔPWC[IGSdaily - ERA5]": df_diff_pwc_daily_era5,
    "ΔPWC[IGSrepro3 - ERA5]": df_diff_pwc_repro3_era5,
    "ΔPWC[EPNrepro2 - ERA5]": df_diff_pwc_repro2_era5,
} 

# Name of the keys are important and must remain the same as shown 
# Adapt the name of your dataframes if you want to use intercomparison with EPNrepro2

diff_columns = {
    "ΔZTD[IGSdaily - IGSrepro3]": "ZTD_diff",
    "ΔZTD[IGSdaily - EPNrepro2]": "ZTD_diff",
    "ΔZTD[IGSrepro3 - EPNrepro2]": "ZTD_diff",

    "ΔPWC[IGSdaily - IGSrepro3]": "PWC_diff",
    "ΔPWC[IGSdaily - EPNrepro2]": "PWC_diff",
    "ΔPWC[IGSrepro3 - EPNrepro2]": "PWC_diff",

    "ΔPWC[IGSdaily - ERA5]": "PWC_diff",
    "ΔPWC[IGSrepro3 - ERA5]": "PWC_diff",
    "ΔPWC[EPNrepro2 - ERA5]": "PWC_diff",
}

station_stats_dict = {}

for label, df in dataframes_dict.items():
    diff_column = diff_columns[label]

    station_stats = df.groupby("Station")[diff_column].agg(["median"]).reset_index()
    station_stats["IQR"] = (
        df.groupby("Station")[diff_column]
        .apply(lambda x: x.quantile(0.75) - x.quantile(0.25))
        .values
    )
    station_stats["mean"] = df.groupby("Station")[diff_column].mean().values
    station_stats["abs_median"] = station_stats["median"].abs()
    station_stats["abs_mean"] = station_stats["mean"].abs()
    station_stats["std"] = df.groupby("Station")[diff_column].std().values
    station_stats["IQR/1.35"] = station_stats["IQR"] / 1.35
    station_stats_dict[label] = station_stats

station_stats_PWC_REPRO3_ERA5 = station_stats_dict["ΔPWC[IGSrepro3 - ERA5]"]
station_stats_PWC_DAILY_ERA5 = station_stats_dict["ΔPWC[IGSdaily - ERA5]"]
station_stats_ZTD_DAILY_REPRO3 = station_stats_dict["ΔZTD[IGSdaily - IGSrepro3]"]
station_stats_PWC_DAILY_REPRO3 = station_stats_dict["ΔPWC[IGSdaily - IGSrepro3]"]

station_stats_ZTD_DAILY_REPRO3_mm = station_stats_ZTD_DAILY_REPRO3.copy()
numeric_columns = ["median", "IQR", "mean", "abs_median", "abs_mean", "std", "IQR/1.35"]
station_stats_ZTD_DAILY_REPRO3_mm[numeric_columns] = (
    station_stats_ZTD_DAILY_REPRO3[numeric_columns] * 1000
)

vf.plot_scatter_joint_stats(
    station_stats_ZTD_DAILY_REPRO3_mm,
    df_equipment,
    dataframes_dict,
    datatype="ZTD",
    x_col="mean",
    y_col="std",
    title="IGSdaily - IGSrepro3",
    xlabel="Mean[ΔZTD] [mm]",
    ylabel="std[ΔZTD] [mm]",
    xlim=(pd.to_datetime("2000-01-01"), pd.to_datetime("2019-12-31")),
    ylim=[-0.1, 0.1],
    annotate_worst=False,
    n_worst=5,
    annotate_threshold_worst=True,
    threshold_mean=(-5, 8),
    threshold_std=13,
    show_grid=True,
    show_stats=True,
    save=False,
    save_path=f"{base_dir}/full_time_series_data/Joint_ZTD_DAILY_REPRO3.png",
    dpi=600,
)

<Figure size 1000x600 with 0 Axes>

<seaborn.axisgrid.JointGrid at 0x72ebad6ea440>

vf.plot_scatter_joint_stats(
    station_stats_PWC_DAILY_REPRO3,
    df_equipment,
    dataframes_dict,
    x_col="mean",
    y_col="std",
    title="IGSdaily - IGSrepro3",
    xlabel="Mean[ΔPWC] [kg/m²]",
    ylabel="std[ΔPWC] [kg/m²]",
    xlim=(pd.to_datetime("2000-01-01"), pd.to_datetime("2019-12-31")),
    ylim=[-10, 10],
    annotate_worst=False,
    n_worst=5,
    annotate_threshold_worst=True,
    threshold_mean=(-0.5, 1.5),
    threshold_std=2.5,
    show_grid=True,
    show_stats=True,
    save=False,
    save_path=f"{base_dir}/full_time_series_data/Joint_PWC_DAILY_REPRO3.png",
    dpi=600,
)

<Figure size 1000x600 with 0 Axes>

<seaborn.axisgrid.JointGrid at 0x72e9156c6fb0>

vf.plot_scatter_joint_stats(
    station_stats_PWC_DAILY_ERA5,
    df_equipment,
    dataframes_dict,
    x_col="mean",
    y_col="std",
    title="IGSdaily - ERA5",
    xlabel="Mean[ΔPWC] [kg/m²]",
    ylabel="std[ΔPWC] [kg/m²]",
    xlim=(pd.to_datetime("2000-01-01"), pd.to_datetime("2023-12-31")),
    annotate_worst=False,
    n_worst=5,
    annotate_threshold_worst=True,
    threshold_mean=(-3, 2),
    threshold_std=3.5,
    show_grid=True,
    show_stats=True,
    save=False,
    save_path=f"{base_dir}/full_time_series_data/Joint_PWC_DAILY_ERA5.png",
    dpi=600,
)

<Figure size 1000x600 with 0 Axes>

<seaborn.axisgrid.JointGrid at 0x72ebe223df30>

vf.plot_scatter_joint_stats(
    station_stats_PWC_REPRO3_ERA5,
    df_equipment,
    dataframes_dict,
    x_col="mean",
    y_col="std",
    title="IGSrepro3 - ERA5",
    xlabel="Mean[ΔPWC] [kg/m²]",
    ylabel="std[ΔPWC] [kg/m²]",
    xlim=(pd.to_datetime("2000-01-01"), pd.to_datetime("2023-12-31")),
    annotate_worst=False,
    n_worst=5,
    annotate_threshold_worst=True,
    threshold_mean=(-3, 2),
    threshold_std=3.5,
    show_grid=True,
    show_stats=True,
    save=False,
    save_path=f"{base_dir}/full_time_series_data/Joint_PWC_REPRO3_ERA5.png",
    dpi=600,
)

<Figure size 1000x600 with 0 Axes>

<seaborn.axisgrid.JointGrid at 0x72ebe2d166b0>

vf.plot_diff_time_series(
    dataframes_dict,
    "MAT100ITA",
    "ZTD_diff",
    labels=["ΔZTD[IGSdaily - IGSrepro3]"],
    equipment_change=df_equipment,
    ylim=None,
    xlim=(pd.to_datetime("01-01-2008"), pd.to_datetime("31-12-2020")),
    save=False,
)

/tmp/ipykernel_583776/4210248729.py:8: UserWarning: Parsing dates in %d-%m-%Y format when dayfirst=False (the default) was specified. Pass `dayfirst=True` or specify a format to silence this warning.
  xlim=(pd.to_datetime("01-01-2008"), pd.to_datetime("31-12-2020")),

vf.plot_diff_time_series(
    dataframes_dict,
    "BAMF00CAN",
    "PWC_diff",
    labels=["ΔPWC[IGSdaily - ERA5]"],
    equipment_change=df_equipment,
    ylim=None,
    xlim=(pd.to_datetime("01-01-2008"), pd.to_datetime("31-12-2023")),
    save=False,
)

/tmp/ipykernel_583776/4134070653.py:8: UserWarning: Parsing dates in %d-%m-%Y format when dayfirst=False (the default) was specified. Pass `dayfirst=True` or specify a format to silence this warning.
  xlim=(pd.to_datetime("01-01-2008"), pd.to_datetime("31-12-2023")),

		observation_id	report_id	observation_value	units	height_of_station_above_sea_level	report_timestamp	city	uncertainty_type1	uncertainty_type5	uncertainty_units1	uncertainty_units5	uncertainty_value1	uncertainty_value5	longitude\|station_configuration	latitude\|station_configuration
primary_station_id	observed_variable
ABMF00GLP	precipitable_water_column	b'16782553'	b'1'	53.900002	b'kg m-2'	15.252470	2008-07-16 00:00:00	b'Les Abymes'	1	5	b'NA'	b'kg m-2'	NaN	0.72	-61.527538	16.262304
	precipitable_water_column	b'16782554'	b'1'	47.369999	b'kg m-2'	15.252470	2008-07-16 01:00:00	b'Les Abymes'	1	5	b'NA'	b'kg m-2'	NaN	0.71	-61.527538	16.262304
	precipitable_water_column	b'16782555'	b'1'	49.630001	b'kg m-2'	15.252470	2008-07-16 02:00:00	b'Les Abymes'	1	5	b'NA'	b'kg m-2'	NaN	0.68	-61.527538	16.262304
	precipitable_water_column	b'16782556'	b'1'	48.529999	b'kg m-2'	15.252470	2008-07-16 03:00:00	b'Les Abymes'	1	5	b'NA'	b'kg m-2'	NaN	0.56	-61.527538	16.262304
	precipitable_water_column	b'16782557'	b'1'	46.389999	b'kg m-2'	15.252470	2008-07-16 04:00:00	b'Les Abymes'	1	5	b'NA'	b'kg m-2'	NaN	0.54	-61.527538	16.262304
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
ZWEN00RUS	precipitable_water_column	b'8003708'	b'422'	8.020000	b'kg m-2'	189.572144	2004-11-06 19:00:00	b'Zwenigorod'	1	5	b'NA'	b'kg m-2'	NaN	0.95	36.758629	55.699284
	precipitable_water_column	b'8003709'	b'422'	7.920000	b'kg m-2'	189.572144	2004-11-06 20:00:00	b'Zwenigorod'	1	5	b'NA'	b'kg m-2'	NaN	0.98	36.758629	55.699284
	precipitable_water_column	b'8003710'	b'422'	9.880000	b'kg m-2'	189.572144	2004-11-06 21:00:00	b'Zwenigorod'	1	5	b'NA'	b'kg m-2'	NaN	0.97	36.758629	55.699284
	precipitable_water_column	b'8003711'	b'422'	8.490000	b'kg m-2'	189.572144	2004-11-06 22:00:00	b'Zwenigorod'	1	5	b'NA'	b'kg m-2'	NaN	0.82	36.758629	55.699284
	precipitable_water_column	b'8003712'	b'422'	11.590000	b'kg m-2'	189.572144	2004-11-06 23:00:00	b'Zwenigorod'	1	5	b'NA'	b'kg m-2'	NaN	0.66	36.758629	55.699284

		observation_id	report_id	observation_value	units	height_of_station_above_sea_level	report_timestamp	city	uncertainty_type1	uncertainty_type5	uncertainty_units1	uncertainty_units5	uncertainty_value1	uncertainty_value5	longitude\|station_configuration	latitude\|station_configuration
primary_station_id	observed_variable
ABMF00GLP	precipitable_water_column	b'21761809'	b'453'	42.410000	b'kg m-2'	15.252470	2010-01-01 00:00:00	b'Les Abymes'	1	5	b'NA'	b'kg m-2'	NaN	0.55	-61.527538	16.262304
	precipitable_water_column	b'21761810'	b'453'	43.099998	b'kg m-2'	15.252470	2010-01-01 01:00:00	b'Les Abymes'	1	5	b'NA'	b'kg m-2'	NaN	0.49	-61.527538	16.262304
	precipitable_water_column	b'21761811'	b'453'	42.299999	b'kg m-2'	15.252470	2010-01-01 02:00:00	b'Les Abymes'	1	5	b'NA'	b'kg m-2'	NaN	0.48	-61.527538	16.262304
	precipitable_water_column	b'21761812'	b'453'	42.889999	b'kg m-2'	15.252470	2010-01-01 03:00:00	b'Les Abymes'	1	5	b'NA'	b'kg m-2'	NaN	0.49	-61.527538	16.262304
	precipitable_water_column	b'21761813'	b'453'	42.290001	b'kg m-2'	15.252470	2010-01-01 04:00:00	b'Les Abymes'	1	5	b'NA'	b'kg m-2'	NaN	0.48	-61.527538	16.262304
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
ZWEN00RUS	precipitable_water_column	b'8312468'	b'184'	15.370000	b'kg m-2'	189.572144	2004-11-03 19:00:00	b'Zwenigorod'	1	5	b'NA'	b'kg m-2'	NaN	0.38	36.758629	55.699284
	precipitable_water_column	b'8312469'	b'184'	15.800000	b'kg m-2'	189.572144	2004-11-03 20:00:00	b'Zwenigorod'	1	5	b'NA'	b'kg m-2'	NaN	0.41	36.758629	55.699284
	precipitable_water_column	b'8312470'	b'184'	14.810000	b'kg m-2'	189.572144	2004-11-03 21:00:00	b'Zwenigorod'	1	5	b'NA'	b'kg m-2'	NaN	0.37	36.758629	55.699284
	precipitable_water_column	b'8312471'	b'184'	14.960000	b'kg m-2'	189.572144	2004-11-03 22:00:00	b'Zwenigorod'	1	5	b'NA'	b'kg m-2'	NaN	0.40	36.758629	55.699284
	precipitable_water_column	b'8312472'	b'184'	12.920000	b'kg m-2'	189.572144	2004-11-03 23:00:00	b'Zwenigorod'	1	5	b'NA'	b'kg m-2'	NaN	0.38	36.758629	55.699284

		observation_id	report_id	observation_value	units	height_of_station_above_sea_level	report_timestamp	city	uncertainty_type1	uncertainty_type5	uncertainty_units1	uncertainty_units5	uncertainty_value1	uncertainty_value5	longitude	latitude
primary_station_id	observed_variable
ACOR00ESP	precipitable_water_column	b'1792297'	b'105'	31.070000	b'kg m-2'	11.931510	1999-08-25 00:30:00	b'A Coruna'	1	5	b'NA'	b'kg m-2'	NaN	1.00	-8.398936	43.364380
	precipitable_water_column	b'1792298'	b'105'	30.120001	b'kg m-2'	11.931510	1999-08-25 01:30:00	b'A Coruna'	1	5	b'NA'	b'kg m-2'	NaN	1.03	-8.398936	43.364380
	precipitable_water_column	b'1792299'	b'105'	29.719999	b'kg m-2'	11.931510	1999-08-25 02:30:00	b'A Coruna'	1	5	b'NA'	b'kg m-2'	NaN	1.09	-8.398936	43.364380
	precipitable_water_column	b'1792300'	b'105'	30.360001	b'kg m-2'	11.931510	1999-08-25 03:30:00	b'A Coruna'	1	5	b'NA'	b'kg m-2'	NaN	0.91	-8.398936	43.364380
	precipitable_water_column	b'1792301'	b'105'	30.500000	b'kg m-2'	11.931510	1999-08-25 04:30:00	b'A Coruna'	1	5	b'NA'	b'kg m-2'	NaN	0.94	-8.398936	43.364380
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
ZYWI00POL	precipitable_water_column	b'27922292'	b'226'	11.900000	b'kg m-2'	370.923523	2014-12-23 19:30:00	b'Zywiec'	1	5	b'NA'	b'kg m-2'	NaN	0.95	19.205954	49.686691
	precipitable_water_column	b'27922293'	b'226'	12.480000	b'kg m-2'	370.923523	2014-12-23 20:30:00	b'Zywiec'	1	5	b'NA'	b'kg m-2'	NaN	1.11	19.205954	49.686691
	precipitable_water_column	b'27922294'	b'226'	12.310000	b'kg m-2'	370.923523	2014-12-23 21:30:00	b'Zywiec'	1	5	b'NA'	b'kg m-2'	NaN	0.95	19.205954	49.686691
	precipitable_water_column	b'27922295'	b'226'	12.410000	b'kg m-2'	370.923523	2014-12-23 22:30:00	b'Zywiec'	1	5	b'NA'	b'kg m-2'	NaN	1.10	19.205954	49.686691
	precipitable_water_column	b'27922296'	b'226'	13.170000	b'kg m-2'	370.923523	2014-12-23 23:30:00	b'Zywiec'	1	5	b'NA'	b'kg m-2'	NaN	1.13	19.205954	49.686691

	Station	Year	Doy	Change_Date	Type
0	abmf	2008	197	2008-07-15 00:00:00	RA
1	abmf	2008	364	2008-12-29 00:00:00	R
2	abmf	2009	288	2009-10-15 00:00:00	RA
3	abmf	2010	239	2010-08-27 00:00:00	R
4	abmf	2011	83	2011-03-24 00:00:00	R
...	...	...	...	...	...
8374	zwen	1996	32	1996-02-01 00:00:00	R
8375	zwen	1999	251	1999-09-08 00:00:00	R
8376	zwen	2000	251	2000-09-07 00:00:00	R
8377	zwen	2000	265	2000-09-21 00:00:00	A
8378	zwen	2004	266	2004-09-22 00:00:00	A

Visualisation and Intercomparison of GNSS In-Situ ZTD and IWV Time Series¶

C3S API Request¶

Concatenation of the yearly files¶

Dataframes¶

IGS REPRO3¶

IGS DAILY¶

EPN REPRO2¶

Overview¶

First Time Series¶

Metadata from Sitelogs IGS files¶

Equipment Changes for IGS Stations¶

Compute differences¶

Compute ΔZTD[IGSDaily - IGSRepro3]¶

Compute ΔIWV [IGSDaily - IGSRepro3]¶

Compute ΔPWC[IGSrepro3 - ERA5]¶

Compute ΔPWC[IGSdaily - ERA5]¶

Visualize the intercomparison of the different datasets¶

Dataframes¶

Get rid of the station with less than 3 years of data¶

Dataframes (only stations with more than 3 years of data)¶

General stats¶

ΔZTD[IGSdaily - IGSrepro3]¶

ΔPWC[IGSdaily - IGSrepro3]¶

ΔPWC[IGSdaily - ERA5]¶

ΔPWC[IGSrepro3 - ERA5]¶

Select your station and plot the time serie¶